Senior Agentic AI Developer

Cynet Systems•Sunnyvale, CA

About The Position

Build reusable agent infrastructure, orchestration capabilities, interfaces, connectors, context services, evaluation, and observability for the organization's AI-Native Core products. The role involves developing agent runtimes and orchestration frameworks, shared AI services and integrations, and reusable platform capabilities for product teams.

Requirements

Agentic AI.
Agent orchestration frameworks.

Responsibilities

Design reusable agent runtimes using LangGraph, Semantic Kernel, AutoGen, CrewAI, and workflow engines.
Develop orchestration frameworks supporting stateful, multi-agent, and event-driven execution patterns.
Create capability registries, tool catalogs, SDKs, and reusable service interfaces.
Build enterprise MCP ecosystems, including servers, clients, gateways, and reusable integrations.
Implement shared context services, memory management, session persistence, and conversation lifecycle controls.
Engineer common RAG services, including ingestion, indexing, embeddings, retrieval, reranking, and grounding.
Develop REST/gRPC APIs for orchestration, capability Client, workflow execution, and agent management.
Build reusable connectors for GitLab, Jira, databases, cloud services, APIs, and enterprise repositories.
Integrate Vertex AI model hosting, routing, quota management, streaming, and cost optimization.
Implement workflow resilience using retries, compensations, fallback strategies, and timeout controls.
Build evaluation frameworks covering task success, safety, latency, reliability, and regression testing.
Implement distributed execution using Pub/Sub, Kafka, queues, schedulers, and asynchronous workflows.
Provide end-to-end observability for models, prompts, workflows, tools, and platform services.
Deploy platform services on GKE and Cloud Run using CI/CD, Infrastructure as Code, and automated release gates.
Deliver reusable reference implementations and onboarding assets consumable across product teams.